Document generated on 10/07/2021 9:29 p.m.

Géographie physique et Quaternaire

Permafrost Distribution in Peatlands of West-Central During the Holocene Warm Period 6000 Years BP La répartition du pergélisol dans les tourbières du centre-ouest du Canada il y a 6000 ans, au cours de la période chaude de l’Holocène Permafrost-Verteilung in Torfmooren von Zentral-West-Kanada während der warmen Holozän-Periode vor 6000 Jahren Stephen C. Zoltai

La paléogéographie et la paléoécologie d’il y a 6000 ans BP au Canada Article abstract Paleogeography and Paleoecology of 6000 yr BP in Canada The extent and distribution of permafrost in peatlands 6000 years ago was Volume 49, Number 1, 1995 investigated in the present discontinuous and continuous permafrost zones of west-central Canada. Permafrost peatlands were cored at 161 locations and the URI: https://id.erudit.org/iderudit/033029ar floristic composition of the peat was determined from macrofossil analysis. DOI: https://doi.org/10.7202/033029ar The reconstructed paleoenvironments were used to indicate the presence or absence of permafrost at the time of peat formation. Chronological control was provided by radiocarbon dating of substantial changes in the peat sequences See table of contents and by dates of basal peat deposits. Peatland formation began after glacial ice disappeared from the land surface of west-central Canada. Macrofossils indicate that most peatlands were fens without permafrost at 6 ka, except in Publisher(s) the . Permafrost was already present in many areas of the , and peat accumulation occurred under permafrost conditions. In the southern Les Presses de l'Université de Montréal area, permafrost development in peatlands began about 4 ka as the middle Holocene warm period came to a close. Permafrost development in the fens ISSN was associated with the development of a Sphagnum-dominated surface on the 0705-7199 (print) fens, caused by the onset of a cooler and moister climate. The insulation 1492-143X (digital) provided by the surface peat layer and by the associated tree cover initiated permafrost development in small lenses that coalesced into large permafrost bodies according to the prevailing climatic conditions. At the target date, 6 ka, Explore this journal permafrost was present in some peatlands, but the distribution zones shifted 300 to 500 km to the north, relative to the present zonation. It is estimated that this corresponds to a mean annual temperature that was about 5°C warmer Cite this article than at present. Zoltai, S. C. (1995). Permafrost Distribution in Peatlands of West-Central Canada During the Holocene Warm Period 6000 Years BP. Géographie physique et Quaternaire, 49(1), 45–54. https://doi.org/10.7202/033029ar

Tous droits réservés © Les Presses de l'Université de Montréal, 1995 This document is protected by copyright law. Use of the services of Érudit (including reproduction) is subject to its terms and conditions, which can be viewed online. https://apropos.erudit.org/en/users/policy-on-use/

This article is disseminated and preserved by Érudit. Érudit is a non-profit inter-university consortium of the Université de Montréal, Université Laval, and the Université du Québec à Montréal. Its mission is to promote and disseminate research. https://www.erudit.org/en/ Géographie physique et Quaternaire, 1995, vol. 49, n° 1, p. 45-54, 3 fig., 4 tabl.

PERMAFROST DISTRIBUTION IN PEATLANDS OF WEST-CENTRAL CANADA DURING THE HOLOCENE WARM PERIOD 6000 YEARS BP

Stephen C. ZOLTAI, Canadian Forest Service, 5320 - 122 Street, Edmonton, T6H 3S5.

ABSTRACT The extent and distribution of RÉSUMÉ La répartition du pergélisol dans ZUSAMMENFASSUNG Permafrost- permafrost in peatlands 6000 years ago was les tourbières du centre-ouest du Canada il Verteilung in Torfmooren von Zentral-West- investigated in the present discontinuous y a 6000 ans, au cours de la période chaude Kanada wàhrend der warmen Holozàn- and continuous permafrost zones of west- de l'Holocène. Des carottes ont été Periode vor 6000 Jahren. Die Ausdehnung central Canada. Permafrost peatlands were prélevées dans 161 sites de tourbières und Verteilung von Permafrost in Torfmooren cored at 161 locations and the floristic com­ pergélisolées et la composition floristique vor 6000 Jahren wurden in den heutigen position of the peat was determined from de la tourbe a été déterminée à partir de diskontinuierlichen und kontinuierlichen macrofossil analysis. The reconstructed l'analyse macrofossile. Les paléo­ Permafrost-Zonen von Zentral-West-Kanada paleoenvironments were used to indicate environnements ainsi reconstitués ont servi untersucht. An 161 Plàtzen hat man Kern- the presence or absence of permafrost at à révéler la présence ou l'absence de bohrungen in den Permafrost-Torfmooren the time of peat formation. Chronological pergélisol au moment de la formation de la durchgefùhrt und mittels der makrofossilen control was provided by radiocarbon dating tourbe. La chronologie a été établie par Analyse die Flora-Kompositon des Torfs of substantial changes in the peat se­ radiodatation des changements majeurs bestimmt. Man erhielt eine chronologische quences and by dates of basal peat depos­ survenus dans les séquences de tourbe et Kontrolle mittels Radiokarbondatierung von its. Peatland formation began after glacial par datation des dépôts de la tourbe basale. gewichtigen Verànderungen in den ice disappeared from the land surface of La formation des tourbières a commencé Torfsequenzen und mittels der Daten von west-central Canada. Macrofossils indicate après la disparition des glaciers de la région. Torfablagerungen an der Basis. Die that most peatlands were fens without per­ Les macrofossiles démontrent que la plupart Torfbildung begann, nachdem das glaziale mafrost at 6 ka, except in the far north. des tourbières étaient de nature Eis von der Landoberflâche von Zentral- Permafrost was already present in many minérotrophe et non pergélisolées à 6 ka, West-Kanada verschwunden war. Die areas of the Arctic, and peat accumulation sauf dans le Grand Nord. Le pergélisol était Makrofossile zeigen, daB die meisten occurred under permafrost conditions. In déjà présent dans beaucoup de secteurs Torfmoore um 6 ka Moore ohne Permafrost the southern area, permafrost development de l'Arctique et l'accumulation de la tourbe waren, auBer im hohen Norden. Permafrost in peatlands began about 4 ka as the mid­ s'est faite concurremment avec la forma­ gab es schon in vielen Gebieten der Arktis, dle Holocene warm period came to a close. tion du pergélisol. Dans la partie und die Torfakkumulation geschah unter Permafrost development in the fens was méridionale, le développement du pergélisol Permafrost-Bedingungen. Im sud- lichen associated with the development of a dans les tourbières a commencé vers 4 ka, Teil begann die Permafrost-Entwicklung in Sphagnum-dominated surface on the fens, à la fin de la période chaude de l'Holocène den Torfmooren um etwa 4 ka, als die caused by the onset of a cooler and moister moyen. Le développement du pergélisol mittlere warme Holozàn-Periode dem Ende climate. The insulation provided by the sur­ dans les tourbières minérotrophes était zuging. Die Permafrost-Entwicklung in den face peat layer and by the associated tree associé à la prédominance de Sphagnum Mooren war mit der Entwicklung einer von cover initiated permafrost development in en surface, suscitée par l'avènement d'un Sphagnum dominierten Oberflàche der small lenses that coalesced into large per­ climat plus froid et plus humide. L'isolation Moore verbunden, hervorgerufen durch den mafrost bodies according to the prevailing procurée par la couche de tourbe en sur­ Beginn eines kùhleren und feuchteren climatic conditions. At the target date, 6 ka, face et par la couverture d'arbres qui lui Klimas. Die Isolierung, die durch die permafrost was present in some peatlands, était associée a entraîné le développement Oberflàchenschicht und durch die damit but the distribution zones shifted 300 to de petites lentilles de pergélisol qui ont verbundene Baumbedeckung hervorgerufen 500 km to the north, relative to the present évolué par coalescence en grands ensem­ wurde, erzeugte Permafrost-Entwicklung in zonation. It is estimated that this corre­ bles en accord avec les conditions kleinen Linsen, welche sich zu breiten Per- sponds to a mean annual temperature that climatiques qui prévalaient. À 6 ka, le mafrost-Einheiten vereinigten, entsprechend was about 5°C warmer than at present. pergélisol était présent dans certaines den vorherrschenden klimatischen tourbières, mais la zone de répartition était Bedingungen. Um 6 ka gab es Permafrost de 300 à 500 km plus au nord que in einigen Torfmooren, aber die maintenant. On estime que la température Verteilungszonen waren 300 bis 500 km devait être de 5°C plus chaude nôrdlicher als heute. Man schàtzt, daB dies qu'actuellement. einer durch-schnittlichen Jahrestemperatur entspricht, die um etwa 5°C warmer als heute war. Manuscrit reçu le 4 février 1994; manuscrit révisé accepté le 25 octobre 1994 46 S.C. ZOLTAI

INTRODUCTION lowed later. The events leading to permafrost aggradation may have been in direct response to a change in climate, or Paleoecological studies and climatic reconstructions it may have been an indirect response to a cooler and showed that the general warm period that followed the moister climate that led to the development of Sphagnum- disappearance of glacial ice from continued dominated peatlands. well into the middle Holocene (COHMAP Members, 1988). This warm, dry period, often referred to as the Hypsithermal Observations in the discontinuous permafrost zone show (altithermal, xerothermic, thermal maximum, Deevey and that permafrost formation is initiated in peatlands mainly as Flint, 1957), began about 10 ka ('000 years before present) particular conditions are created by the vegetation. The in and peaked from 9 to 6 ka in the Great most common sequence is where small Sphagnum cush­ ions develop on the surface of fens (Zoltai and Tarnocai, Plains (Anderson et al., 1989). In the continental parts of 1975). These cushions, projecting above the water table, North America and the summer temperatures at 6 insulate the seasonal frost against rapid summer thawing. ka were 2 to 4°C higher than at present (COHMAP Mem­ Eventually a small lens of frost fails to thaw under the bers, 1988). These temperatures influenced the formation cushion and becomes permafrost. As this frost lens thick­ of postglacial peatlands and the development of permafrost ens, the water in the peat changes to ice, further elevating in the peatlands. the surface above the water table and increasing its insula­ The present distribution of permafrost shows a latitudinal tion value. Small trees often become established on the zonation where the incidence and thickness of permafrost cushions, intercepting some of the snow and thus reducing increases with latitude. Although there were numerous at­ the insulation effect of the snow layer. This process, which tempts at delineating and naming these zones, no consen­ eventually reaches an equilibrium with the prevailing cli­ sus has been reached (Nelson, 1989). The following zones mate, appears to be the main mechanism of permafrost are described in terms of permafrost development in initiation in the SPZ and WPZ. peatlands (Fig. 1). At the most southerly extent, permafrost occurs as small, isolated lenses in peat (Zoltai, 1972) that A second mechanism for the initiation of permafrost is also prevalent, mainly in the LPZ (Zoltai, 1972). Here local reach into the mineral substrate only under palsas. This has patches (several square metres) of dense tree growth can been named the Localized Permafrost Zone (LPZ, Zoltai, develop in large bogs, or as small islands in fens. The 1971). Farther north the permafrost lenses occur more fre­ dense tree canopy (Picea mariana) intercepts much of the quently in bogs or as isolated islands in minerotrophic fens, snow, resulting in a much reduced snow cover under the but unfrozen peatlands are still more extensive than trees (Zoltai and Tarnocai, 1971). Without a thick insulating peatlands with permafrost (Sporadic Permafrost Zone, SPZ). snow blanket, increased heat loss occurs during the winter, Here, permafrost usually reaches the mineral soil beneath causing frost to penetrate deeper under the dense tree both the peat plateaus and palsas. Still farther north in the cover. Some of the seasonal frost may fail to thaw com­ Widespread Permafrost Zone, (WPZ), most peatlands con­ pletely during the summer and hence becomes permafrost. tain permafrost, but many fens and all shallow lakes are Initially the permafrost consists only of thin frozen layers, free of permafrost. The most northerly is within the but these can consolidate into larger lenses as the surface Continuous Permafrost Zone, (CPZ) where all land sur­ is elevated by the freezing of the water (Zoltai, 1972). This faces and even shallow water bodies are underlain by per­ in turn creates drier surface conditions, further promoting mafrost. dense tree growth and less snow cover on the surface, The generally parallel latitudinal zonation suggests a enhancing further permafrost accretion. relationship to climate. The mean annual temperature in the Isolated permafrost bodies have been observed in LPZ is between 0 and -1°C (Vitt et al., 1994), and in the SPZ it is between -1 and -5.5°C (Brown, 1978). In the wetlands that have only a thin peat layer (Brown, 1980; WPZ the mean annual temperature is generally between Seppâlâ, 1980; Tarnocai, 1982), and their internal -5.5 and -8.3°C (Nelson, 1989); in the CPZ the mean stratigraphy does not suggest any relationship to peatland annual temperature is lower than -8.3°C. It is recognized, dynamics. No mechanism has been suggested for their however, that the development of permafrost in peatlands formation, but one can speculate that local frost heaving in is also influenced by other factors, such as timing and the underlying mineral soil can raise the surface sufficiently amount of snowfall, annual precipitation and the presence above the water table to allow the desiccation of the peat of peatlands that are susceptible to permafrost develop­ layer, hence creating an insulating layer (Seppâlâ, 1988). ment. Various permafrost mounds (Outcalt and Nelson, 1984; Pollard and French, 1983) and frost blisters (Van Everdingen, Previous research has established the dynamics of per­ 1982) that may be present in some wet areas are formed by mafrost development in peatlands (Zoltai, 1972; Zoltai and the injection of ice between the permafrost table and the Tarnocai, 1975; Payette et ai, 1976; Couillard and Payette, freezing surface layer. The genesis of these forms is, how­ 1985; Seppâlâ, 1988). It was found that in the present LPZ, SPZ and WPZ areas, most of the permafrost developed in ever, not relevant to permafrost initiation and development non-frozen peatlands. Macrofossil analyses of the peat in peatlands, and they are not considered here. showed that most of the peat was deposited in a non- In this paper the extent and timing of permafrost devel­ permafrost environment, and permafrost development fol­ opment in peatlands of west-central Canada are examined.

Géographie physique et Quaternaire. 49(1), 1995 WEST-CENTRAL CANADA 47

FIGURE 1. Permafrost zones at present, based on permafrost dis­ tribution in peatlands, with loca­ tions of peatland study sites. Zones de pergélisol actuelles, à partir de la répartition du pergélisol dans les tourbières et localisation des sites de tourbières étudiées.

PRESENT PERMAFROST ZONES CPZ-N -continuous-north SURFACE CPZ-S -continuous-south lowland polygons WPZ -widespread polygonal peat plateaus SPZ -sporadic peat plateaus LPZ -localized palsas

An estimate of permafrost distribution in peatlands during LPZ, 30 sites have been examined, mainly in the Holocene warm period is presented for the target date and . In the SPZ, 26 sites were located, in Mani­ of 6 ka. toba, Alberta, and the . The WPZ was STUDY AREA represented by 71 sites (98 cores), mainly from the Macken­ zie Valley in the Northwest Territories, but also including The data base consists of field information obtained by sites from , Alberta, and Manitoba. In the CPZ, 34 coring and sampling of peatlands in west-central Canada sites (52 cores) were located. The permafrost peat landforms during the past 23 years. Parts of this information have examined were palsas (30) and peat plateaus (95) in the been published (Zoltai, 1972; Zoltai and Tarnocai, 1975; LPZ, SPZ and WPZ, and polygonal peat plateaus (13 sites), Zoltai, 1993), although much is still unpublished. In the low centre (3) and high centre polygons (16).

Géographie physique el Quaternaire, 49(1), 1995 48 S.C. ZOLTAI

METHODS feather moss remnants (mainly Hypnum sp., Pleurozium sp., Hylocomium sp.) can be identified. Coring sites were selected generally on the highest parts of palsas, or at the centres of peat plateaus. In large peat FEN PEATLANDS plateaus, the coring locations were at least 20 m from the Sedge-moss peat — Peat composed of remnants (roots, edge of the plateau. At the selected sites small excavations leaves) of Carex spp. and brown mosses (Drepanocladus were made to sample the peat of the active layer. The spp., Campylium spp., Scorpidium sp., often with vari­ permafrost was sampled with a small diameter permafrost ous amounts of fragments of wood, twiglets and woody probe (Zoltai, 1978) until mineral soil was reached, or until rootlets. The peat is moderately to well decomposed. at least one metre of clear ice was penetrated. Whenever possible, the adjacent unfrozen peat was cored with a Humic peat — Peat consisting of well-decomposed organic Macauley peat sampler; samples were taken at 10-cm in­ matter, where the constituents are usually no longer tervals. At some locations where peat sections were ex­ recognizable. posed by erosion, the peat face was cut back at least 15 cm to expose uncontaminated materials and the regular sam­ RESULTS pling procedure was followed. The results of the analysis show that nearly all perma­ In the laboratory, the samples were examined to deter­ frost peatlands are covered by a thin surface layer of peat mine the main components of the peat. When possible, that differs in origin from the underlying fen peat (Table I). distinctions were made between Sphagnum species grow­ This surface layer is the thickest in the south and de­ ing in wet and drier habitats. The samples were further creases gradually towards the north. analysed to determine their chemical and physical charac­ teristics. The surface peat material varies among the various per­ mafrost zones, according to the regional vegetation devel­ The peat samples were grouped on the basis of their opment of the permafrost peatlands. Well-decomposed sylvic composition into the following general groups indicating peat is prevalent in the LPZ under densely forested peat broad environmental conditions. plateaus and palsas. In the SPZ where open canopied SPH/AGA/L/M-DOMINATED BOGS AND POOR FEN black spruce with Sphagnum fuscum, some ground lichens PEATLANDS and Labrador tea (Ledum groenlandicum) form the vegeta­ tion cover; Sphagnum or woody Sphagnum peat is most Sphagnum peat — Peat composed dominantly of Sphag­ common. In the WPZ the same species are present, but num fuscum remains, but often including S. magellanicum with fewer, stunted trees. The ground cover is lichen (mainly and S. angustifolium. Ericaceous shrub twiglets and roots Cladina mitis, C. rangiferina, C. stellaris) and Labrador tea, are often present. Conifer (mainly Picea mariana) nee­ with Sphagnum mosses in moister depressions. The lichen dles, cones, or twiglets may be present. This peat is is usually not represented in the peat, leaving ericaceous generally well preserved because it usually occurs near shrub debris and Sphagnum remains as the main peat the surface. constituents. Woody Sphagnum peat — Peat consisting mainly of Along the southern fringe of the CPZ, (CPZ-S) the ericaceous shrub stems, roots and leaves, with variable peatlands are covered with lichen (Cetraria nivalis, C. amounts of Sphagnum fuscum, and rarely, lichen {Cladina cucullata, Alectoria spp.), with Sphagnum fuscum in the sp.) remains. The peat is slightly to moderately decom­ moister areas. In this area the surface Sphagnum layer is posed. underlain by more Sphagnum peat at six of the sites. In such cases the surface layer was determined by the greater FORESTED PEATLANDS degree of decomposition observed, compared to that in the Sylvie (forest) peat — Peat consisting of well decomposed underlying Sphagnum peat. Nearly all peatlands occur in woody material, rootlets, and needles. On occasion lowland polygons (Zoltai and Tamocai, 1975), either in low-

TABLE I

Number of sites with different surface materials, and the average thickness of surface layer above fen peat (with standard error) in various permafrost zones

Zone Sites/Cores Sylvic Sphagnum Woody Sedge- Humic Sphagnum moss cm n cm n n

LPZ 30/30 27 55.8±3.1 2 80.0±14.1 SPZ 27/27 5 50.6±2.3 21 56.5±4.6 WPZ 71/98 5 51.6±6.4 65 42.7±5.3 22 33.0±7.0 CPZ-S 14/29 1 39 28 44.9±8.4 CPZ-N 20/23 - - 5 43.0±10.7 18 30.5±1.7

Géographie physique et Quaternaire. 49(1), 1995 WEST-CENTRAL CANADA 49

centre forms with wet centres or in high-centre forms which Only those dates were included that were obtained from have domed, well drained centres. Low-centre polygons peat of fen deposits; those from lacustrine organic sediments are common in the northern CPZ (CPZ-N), with only thin were not included. In addition, basal dates from well-de­ peat development. Here, most peaty high-centre polygons composed (humic) peat of uncertain origin often found in have a bare surface, where the peat is oxidizing. The sur­ the Arctic were included. The basal dates are shown in face peat usually consists of highly humified organic mate­ Figure 2 as being older or younger than 6 ka. rial. Radiocarbon dates have been obtained from the top of In most permafrost zones there are only a very few the fen peat sequence, immediately below the surface permafrost peatlands with fen peat at the surface. These Sphagnum layer, from permafrost peatlands in the discon­ are generally on the margins of peat plateaus and were tinuous (WPZ, SPZ and LPZ) permafrost zones (Table II). probably affected by permafrost as the main permafrost These dates show that the earliest development of the body developed. Only in two peatlands did the permafrost Sphagnum cap was about 3.7 ka, and that it continued appear to have originated in the fens. since then at various locations. The dates appear to indi­ cate that the Sphagnum cap developed earlier in the SPZ CHRONOLOGY than in the more northerly WPZ. Published radiocarbon dates of basal peat deposits were In the southern fringe of the CPZ the transition between collected from 117 locations within the permafrost zones. the decomposed surface layer and the underlying Sphag-

BASAL PEAT DATES PRESENT PERMAFROST ZONES

• greater than 6 ka CPZ-N -continuous -north o less than 6 ka CPZ-S -continuous -south WPZ -widespread ^fO SPZ -sporadic ^C? X? LPZ -localized ^t,0

FIGURE 2. Basal peat dates in the present permafrost zones, showing older than 6 ka and younger that 6 ka dates. Datations de tourbe basale dans les zones de pergélisol actuelles selon qu'elles sont plus vieilles ou plus jeunes que 6 ka.

Géographie physique el Quaternaire, 49(1), 1995 50 S. C. ZOLTAI

num or fen peat was dated between 5.8 and 0.5 ka (Table DISCUSSION III), a similar spread of dates to that in the discontinuous The trend of regional development of peatlands in the permafrost zones. The dates of the thicker Sphagnum de­ continental boreal region (including the LPZ, SPZ, and WPZ) posits indicate that bog conditions were established be­ shows that initial peatland development proceeds to form tween 6.6 and 4.7 ka in this area. fens on mineral soil in poorly drained areas (paludification) In the northern CPZ there appear to be several groups of or over ponds (terrestrialization). At a later stage, Sphag- peatland dates. One group shows basal dates in the 6.0 to nivm-dominated fens may be established in parts of the 10.1 ka range (Table IV, Nos. 1 to 5). The surface dates of fens that may proceed to develop into bogs when the sur­ these peatlands are also old, in the 5.0 to 8.0 ka range. face is no longer influenced by the water table of the fen Other basal dates are younger (Table IV, Nos. 10 to 16), (Zoltai and Johnson, 1985; Nicholson and Vitt, 1990; Kuhry dominantly in the 4.0 to 5.0 ka range with relatively recent et al., 1992). In addition to local water quality conditions, surface dates, in the 2.0 to 3.0 ka range. The remaining the development of bogs is limited by climate. Few bogs dates are intermediate (Table IV, Nos. 6 to 9), with relatively are formed if the summer precipitation is lower than 250 old basal dates (6.0 to 8.0 ka), although the surface peat is mm and the mean annual temperature is above 20C (Vitt et in the 1.0 to 3.0 ka range. al., 1994).

TABLE Il

Radiocarbon dates at the change from fen peat to Sphagnum peat in the discontinuous permafrost zones (WPZ. SPZ. LPZ)

Zone Location Depth (cm) Date Lab No. Source

1. WPZ 68°26'N 133°52'W 27-29 530±120 AECV 974C This paper 2. WPZ 67°30'N 133°46'W 41-46 670±100 AECV 972C This paper 3. WPZ 67°28'N 134°40'W 42-46 1980±200 AECV 976C This paper 4. WPZ 67°28'N 134°35'W 98-104 1720±100 AECV 977C This paper 5. WPZ 67°25'N 134° 18' W 59-62 1660±110 AECV 975C This paper 6. SPZ 61°28'N 120°55'W 155 3000 WAT-400 Chatwin 1983 7. SPZ 60°03'N 112°54'W 67-70 2870±90 AECV 993C This paper 8. SPZ 59°07'N 118°08'W 109-113 2410±120 AECV 985C Zoltai, 1993 9. SPZ 58°18'N 119°17'W 83-86 2710±100 AECV 990C Zoltai, 1993 10. LPZ 54°36'N 101°26'W 84-90 1200±100 BGS 863 This paper

TABLE III

Radiocarbon dates of peat from polygonal peat plateaus of the southern continuous permafrost zone (CPZ-S)

Location Depth (cm) Stratigraphy Date Lab. No. Source

1. 68°22'N 132°42'W 50-60 Sp/iagnum/wood 5420±70 WIS-279 Dyck & Fyles, 1963 360-366 wood/mineral 8200±300 GSC-25 Dyck & Fyles, 1963 2. 67°49'N 139°50'W 21.5-24.5 humic/ Sphagnum 3025±85 S-1865 Ovenden,1982 175-178 sedge/lacustrine 10080±340 S-1871 Ovenden, 1982 3. 67°41'N 132°05'W 27-32 Sphagnum/sedge, moss 2710±60 BGS 147 Deiorme & Zoltai, 1984 266-269 sedge, moss/min* 8610±100 BGS 149 Delorme & Zoltai, 1984 4. 67°06'N 125°47'W 34-36 humic/Sphagnum 1810±60 WIS-297 Nichols, 1974 174-176 Sphagnum/sedge 6630±85 WIS-299 Nichols, 1974 5. 63°04'N 110°47'W 24-26 humic/Sphagnum 3380±105 GaK-5050 Nichols, 1975 160-163 sedge,moss/min 6080±150 BGS 474 This paper 157-159 Sphagnum 6170±130 GSC-1840 Nichols, 1975 6. 63°01'N 128°48'W 30 peat 3000±50 GSC-3176 MacDonald, 1983 230 humic 8640±160 GSC-3097 MacDonald, 1983 7. 61°15'N 100°57'W 4-6 hum\c/Sphagnum 870±60 GaK-5062 Nichols, 1975 288-290 Sphagnum 4690±140 GSC-1781 Nichols, 1975 8. 61°15'N 100°55'W 20-22 Sphagnum 1510±80 WIS-88 Bender et a/., 1966 148-150 Sphagnum 5780±110 WIS-67 Bender et al.. 1966 9. 61°10'N 100°55'W 4-6 Sphagnum 630±70 WIS-133 Nichols, 1967 108-110 Sphagnum 4800±90 WIS-166 Nichols, 1967 10. 61°08'N 97°06'W 13-15 humic/Sphagnum 1920±80 BGS-478 This paper 135-138 sedge, moss/min 5110±130 BGS 477 This paper 11. 60°37'N 96°59'W 20-25 humic/ Sphagnum 550±80 BGS 475 This paper

*min = mineral soil

Géographie physique el Quaternaire, 49(1). 1995 WEST-CENTRAL CANADA 51

TABLE IV Radiocarbon dates of peat from high centre peat polygonsin the northern continuous permafrost zone (CPZ-N)

Location Depth (cm) Stratigraphy Date Lab No. Source

1. 75°45'N 98°18'W 9-12 humic/peat 1170±150 GSC-402 Lowdon et ai, 1967 312-326 peat/ice 6510±150 GSC-253 Lowdon era/., 1967 2. 75°40'N 97°40'W 25 humic peat 5070±60 GSC-2326 Tarnocai & Zoltai, 1988 130 peat 6160±90 GSC-2317 Tamocai & Zoltai, 1988 3. 73°16'N 118°50'W 15 humic/moss 2800±100 BGS 698 Tarnocai & Zoltai, 1988 178 moss/mineral 4250±100 BGS 699 Tarnocai & Zoltai, 1988 4. 72°58'N 94°57'W 30-32 humic/sedge, moss 6070±100 BGS 337 This paper 133-135 sedge, moss/min" 6280±80 GSC-2339 This paper 5. 72°52'N 93°37'W 0-3 humic 1320±60 GSC-2945 Blake, 1987 225-230 peat/mineral 7590±80 GSC-2583 Blake, 1987 6. 71°56'N 123°14'W 61 humic 6940±110 GSC-10 Dyck & Fyles, 1963 244 peat/mineral 9820±220 GSC-197 Dyck & Fyles, 1963 7. 70°07'N 93°33'W 5-8 humic 2080±60 GSC-3282 Blake, 1988 130-132 peat/mineral 4750±60 GSC-3277 Blake, 1988 8. 69°25'N 131°40'W 15 humic/moss 2920±130 GSC-1669 Lowdon & Blake, 1981 80 moss 6770±140 GSC-1737 Lowdon & Blake, 1981 9. 69°15'N 138°02'W 38-40 humic/sedge, moss 8260±110 BGS 196 This paper 298-300 sedge, moss/min 10100±130 BGS 197 This paper 10. 69°09'N 134°17'W 58 peat 1890±60 Beta-11569 Tarnocai & Zoltai, 1988 310 peat 4810±60 Beta-11564 Tarnocai & Zoltai, 1988 11. 69°07'N 123°56'W 20-22 humic/sedge, moss 3150±90 BGS 216 Tarnocai & Zoltai, 1988 320-324 sedge, moss/min 6020±100 BGS 217 Tarnocai & Zoltai, 1988 12. 68°45'N 120°39'W 5-10 humic/peat 2200±80 GSC-5200 Zoltai et al., 1992 115-117 meat/mineral 4030±100 GSC-5188 Zoltai et al., 1992 13. 63°39'N 95°50'W 20-22 humic/sedge, moss 2130±80 BGS 404 This paper 150-152 sedge, moss/min 3850±400 BGS 406 This paper 14. 62°07'N 96°32'W 20-24 humic/sedge, moss 920±80 BGS 484 This paper 202-204 sedge, moss/min 3890±160 BGS 483 This paper 15. 61°38'N 97°12'W 23-25 humic/sedge, moss 2300±100 BGS 481 This paper 174-177 sedge, moss/min 4380±130 BGS 480 This paper

* min - mineral soil

Radiocarbon dates of basal peat show that peat deposits Alberta, bog development occurred earlier in the north (about were widespread throughout the area at 6 ka (Fig. 2). Excep­ 5 ka) than farther south in the present LPZ (about 2.2 ka, tions occurred along the southern boundary of the LPZ, Zoltai and Vitt, 1990). where all ages are <6 ka. This observation has been attrib­ In the CPZ-N, the available data on the age of the uted to lack of peatland development during an arid and surface layer appear to be contradictory. Under the present warm climatic period prior to 6 ka (Zoltai and Vitt, 1990). A climatic regime, limited peat formation takes place in low- similar grouping of <6 ka peatlands occurs in the CPZ-S, centre polygons, where the wetland is underlain by perma­ west of (Fig. 2), in an area that became free of frost. Here the peaty material generally shows the effects of glacial ice cover between 7 and 8 ka ago (Dyke and Prest, mixing by frost action (cryoturbation) as evidenced by the 1987). In general, all peatlands are <6 ka on the eastern presence of mineral sand grains or stones in the peat, mainland, generally east of Great Bear and Great Slave evident in both low and high centre polygons. The peat lakes, which became deglaciated later than the west. contains large amounts of mineral soil particles (35 to 50% Permafrost development in peatlands is associated with by weight), as well as small stones and boulders (Zoltai er 3 the development of bogs (SPZ and WPZ) or the initiation of a/., 1992). Boulders up to 0.02 m can be found embedded Sphagnum-nch treed fens (LPD). In much of the discontinu­ in the peat or on the surface underlain by peat on high ous permafrost zone, however, bog or treed fen develop­ centre polygons where the basin configuration excludes ment did not take place until after 5 ka, indicating that fluvial deposition. Frost heaving and mixing as the peat is permafrost development was also lacking prior to this date. being deposited is a likely mechanism for the distribution of Basal dates indicate that Sphagnum development and the mineral soil grains and stones in the peat. Such evidence of associated permafrost development in peatlands began in cryoturbation was seen in high-centre polygons at nine of the CPZ-S about 6 ka. There are indications that bog devel­ the 20 sites cored, as well as in the shallow peat of low opment and associated permafrost formation began earlier centre polygons, indicating that these peats were deposited in a permafrost environment. in the north and gradually spread southward. Within

Géographie physique el Quaternaire. 49(1), 1995 52 S.C. ZOLTAI

In some areas in the CPZ-N, however, peat formation 2. About 6 ka the dominant form of peatland in the present can occur in the absence of permafrost, as shown by the SPZ and southern WPZ was fen, which covered a much recent dates of surface peat. A number of high centre smaller portion of the area than at present. Permafrost was polygons are composed of old peat, deposited before 6000 absent from these fens. BP in non-permafrost environments. The absence of a suf­ 3. In the northern part of the present WPZ, small Sphag- ficient number of reliably dated peat sequences may well ntvm-dominated fens developed, with small permafrost lenses be the source of this apparent contradiction. in some of these, similar to those found in the present LPZ. 4. In the present CPZ-S, bog development was well ad­ RECONSTRUCTION OF PALEOCLIMATE vanced and scattered permafrost was present in some of them, resembling the permafrost distribution in the southern Based on the available evidence, the following general SPZ of the present. scenario of permafrost distribution in the peatlands at about 5. In the western part of the present CPZ-N a major peat- 6 ka can be generated: 1 forming period came to a close at about 6 ka. In the south­ 1. In the approximate area of the present LPZ, the climate ern portion of this western area permafrost occurred in a was warm and dry. No peatlands and no permafrost oc­ large number of peatlands, in proportions that are possibly curred in this area at 6 ka. similar to the WPZ of the present.

6 ka PERMAFROST ZONES

CPZ -continuous WPZ -widespread SPZ -sporadic LPZ -localized

FIGURE 3. Approximate posi­ tion of permafrost zones at 6 ka, based on permafrost distribution in peatlands. Emplacement approximatif des zones de pergélisol à 6 ka, à partir de la répartition du pergélisol dans les tourbières.

Géographie physique el Quaternaire, 49(1), 1995 WEST-CENTRAL CANADA 53

6. In the mainland part of the present CPZ-N, especially CONCLUSIONS near Hudson Bay, peat formation did not take place at this This paper demonstrates that a knowledge base of time. wetland development processes exists to make an ecologi­ 7. In the northern portion of the present CPZ peat devel­ cally sound estimate of the paleoclimate of permafrost dy­ opment occurred under permafrost conditions, but the namics at 6 ka. However, the accuracy and reliability of this peatlands were limited in extent. reconstruction is questionable because of the uneven and This scenario permits the tentative reconstruction of per­ incomplete data base. With the addition of carefully se­ mafrost distribution in the peatlands of west-central Canada lected sites, much of the uncertainty could be removed and about 6 ka (Fig. 3). This map must be regarded as an initial the paleoclimatic conditions responsible for the distribution reconstruction because of the limited number of sites and of permafrost in peatlands could be determined at a conti­ their uneven distribution. nental scale from the study of peatlands. This would pro­ vide an independent test of paleoclimate reconstructions Further constraints are posed by the allogenic proc­ obtained from the study of vegetation or faunal ecosys­ esses that create conditions under which permafrost can be tems. generated. First, appropriate water quality and climatic con­ ditions must be present before Sphagnum-dominated peatlands can form. Next, the local thermal conditions of ACKNOWLEDGEMENTS the prevailing climate would have to be modified by the I wish to thank both reviewers Hugh French and Inez Sphagnum-dominated cover to initiate permafrost develop­ Kettles for their useful suggestions. ment. Furthermore, there is instrumented evidence of a current warming trend (1.4°C per century) in central West­ REFERENCES ern Canada (Gullett, 1992). The possibility exists that the present permafrost distribution in the peatlands indicates a Anderson, T.W., Mathewes, R.W. and Schweger, CE., 1989. Holocene cooler past climate than the present. Permafrost instability climatic trends in Canada with special reference to the Hypsithermal interval, p. 520-528. In R.J. Fulton, éd., Quaternary Geology of Canada in the LPZ and SPZ may be an indication that permafrost and . Geological Survey of Canada, No. 1. conditions are presently undergoing adjustment to a warmer Bender, M.M., Bryson, R.A. and Baerreis, DA., 1966. University of Wiscon­ climate. sin radiocarbon dates II. Radiocarbon, 8:522-533. The tentative map of permafrost distribution in peatlands Blake. W. Jr., 1987. Geological Survey of Canada radiocarbon dates XXVI. at 6 ka shows a general displacement of the zones as Geological Survey of Canada Paper 86-7, 60 p. compared to the present (Fig. 3). In particular, the LPZ was 1988. Geological Survey of Canada radiocarbon dates XXVII. Geo­ much farther north and more extensive than at present, logical Survey of Canada Paper, 87-7, 100 p. approximately covering the southern half of the present Brown, G., 1980. Palsas and other permafrost features in the lower Rock area of the WPZ. The northward displacement of the CPZ Creek Valley, west-central Alberta. Arctic and Alpine Research, 12:31- was greater in the west than in the area close to Hudson 40. Bay. Brown, R.J.E., 1978. Permafrost map of Canada. Plate 32. In Hydrological Atlas of Canada. Department of Fisheries and Environment, . Some cautious paleoclimate implications may be drawn Chatwin, S.C1 1983. Holocene temperatures in the upper Mackenzie Valley from the reconstructed 6 ka permafrost distribution. The determined by oxygen isotope analysis of peat cellulose, p. 127-130. In mean annual temperature at the southern boundary of the Proceedings, Fourth International Conference on Permafrost. National present LPZ is at 00C, but the corresponding boundary at 6 Academy Press, Washington, D.C ka was at about the current -5°C isotherm. Similarly, the COHMAP Members, 1988. Climatic changes of the last 18,000 years: present boundary between the SPZ and WPZ is at about Observations and model simulations. Science, 241:1043-1052. -5.5°C whereas the 6 ka boundary was at about -10 to Couillard, L. and Payette, S., 1985. Évolution holocène d'une tourbière à -110C. The southern boundary of the CPZ is close to the pergélisol (Québec nordique). Canadian Journal of Botany, 63:1104- -8.3°C isotherm, but it was probably near the present 1121. -12°C isotherm at 6 ka. In general, the northward shift Deevey, E.S. and Flint, R.F., 1957. Postglacial Hypsithermal interval. Sci­ indicates that the mean annual temperatures were about ence, 125: 182-184. 50C warmer than at present. Delorme, L.D. and Zoltai, S.C, 1984. Distribution of an arctic ostracod fauna in space and time. Quaternary Research, 21: 65-73. A comparison of the 6 ka permafrost distribution with Dyck, W. and Fyles, J.G., 1963. Geological Survey of Canada radiocarbon projected permafrost boundaries under equilibrium condi­ dates I and II. Geological Survey of Canada Paper, 63-21, 31 p. tions resulting from a postulated surface temperature change Dyke, AS. and Prest, V.K., 1987. Paleogeography of northern North America of 4 to 5°C due to greenhouse climate warming (Woo et al., 18,000-5,000 years ago. Geological Survey of Canada. Map 1703A, 1992) shows comparable northward shifts. The position of scale 1:12,500,000. the southern limit of permafrost at 6 ka was farther north in Gullett, D.W., 1992. Development of an historical Canadian climate data­ the west than in the greenhouse warming projection, al­ base for temperature, p. 27-31. In GW. Kite and D.K. Harvey, eds., Using Hydrometric Data to Detect and Monitor Climatic Change. Pro­ though the boundaries nearly coincide east of Great Slave ceedings National Hydrology Research Institute Workshop No. 8. Lake. The position of the southern boundary of the CPZ is Kuhry, P., Nicholson, B.J., Gignac, L.D., Vitt, D.H. and Bayley, SE.. 1992. nearly identical in both the greenhouse warming and the Development of Sprtagnum-dominated peatlands in boreal continental 6 ka projections. Canada. Canadian Journal of Botany, 71:10-22.

Géographie physique et Quaternaire, 49(1), 1995 54 S.C. ZOLTAI

Lowdon, J.A. and Blake, W., Jr., 1981. Geological Survey of Canada 1988. Palsas and related forms, p. 247-278. In M.J. Clark, éd., radiocarbon dates XXI. Geological Survey of Canada Paper 81-7, 22 p. Advances in periglacial geomorphology. John Wiley, Chichester. Lowdon, J.A., Fyles, J.G. and Blake, W. Jr., 1967. Geological Survey of Tarnocai, C, 1982. Soil and terrain development on the York Factory Canada radiocarbon dates Vl. Geological Survey of Canada Paper 67- peninsula, Hudson Bay Lowland. Naturaliste canadien, 109:511-522. 2, 42 p. Tarnocai, C. and Zoltai, S.C, 1988. Wetlands of Arctic Canada, p. 27-53. In MacDonald, G.M., 1983. Holocene vegetation history of the upper Natla C.D.A. Rubec, coord., Wetlands of Canada. Polyscience Publications, River area, Northwest Territories, Canada. Arctic and Alpine Research, Montréal. 15:169-180. Van Everdingen, R.O., 1982. Frost blisters of the Bear Rock Spring area near Fort Norman, N.WT. Arctic, 35:243-265. Nelson, F.E., 1989. Permafrost zonation in : A review of published maps. Physical Geography, 10:233-248. Vitt, D.H., Halsey, L.A. and Zoltai, S.C, 1994. Bog landforms of continental western Canada in relation to climate and permafrost patterns. Arctic Nichols, H., 1967. Postglacial history of vegetation and climate at Ennadai and Alpine Research, 26:1-13. Lake, Keewatin and Lynn Lake, Manitoba. Eiszeitalter und Gegenwart, 18:176-197. Woo, M.-k., Lewkowicz, A.G. and Rouse, W.R., 1992. Response of the Canadian permafrost environment to climatic change. Physical Geogra­ 1974. Arctic North American paleoecology: The recent history of phy, 13: 287-317. vegetation and climate deduced from pollen analysis, p. 637-667. In J.D. Ives and R.G. Barry, eds., Arctic and Alpine Environments. Methuen, Zoltai, S.C, 1971. Southern limit of permafrost features in peat landforms, London. Manitoba and Saskatchewan. Geological Association of Canada, Spe­ cial Paper No. 9: 305-310. 1975. Palynological and paleoclimatic study of the Late Quaternary 1972. Palsas and peat plateaus in central Manitoba and Saskatch­ displacement of the boreal forest - tundra ecotone in Keewatin and ewan. Canadian Journal of Forest Research, 2:291-302. Mackenzie, N.WT, Canada. Institute of Arctic and Alpine Research, Occasional Paper 15, 87 p. 1978. A portable sampler for perennially frozen stone-free soils. Canadian Journal of Soil Science, 58:521-523. Nicholson, BJ. and Vitt, D.H., 1990. The paleoecology of a peatland complex in continental western Canada. Canadian Journal of Botany, 1993. Cyclic development of permafrost in the peatlands of north­ 68:121-138. western Alberta. Arctic and Alpine Research, 25:240-246. Zoltai, S.C. and Johnson, J.D., 1985. Development of a treed bog island in Outcalt, S.I. and Nelson, F. 1984. Growth mechanisms in aggradation a minerotrophic fen. Canadian Journal of Botany, 63:1076-1085. palsas. Zeitschrift fur Gletscherkunde und Glazialgeologie, 20:65-78. Zoltai, S.C, Sirois, J. and Scotter, G.W., 1992. A natural resource survey Ovenden, L., 1982. Vegetation history of a polygonal peatland, northern of the Melville Hills region, Northwest Territories. Canadian Wildlife Yukon (Canada). Boreas, 11:209-224. Service, Western and Northern Region, Technical Report Series No. Pollard, W.H. and French, H.M., 1983. Seasonal frost mound occurrence, 135, 121 p. North Fork Pass, Ogilvie Mountains, Northern Yukon, Canada. In Pro­ Zoltai, SC and Tarnocai, C, 1971. Properties of a wooded palsa in ceedings, Fourth International Conference on Permafrost, 1:1000-1004. northern Manitoba. Arctic and Alpine Research, 3:115-129. Payette, S., Samson, H. and Lagarec, D., 1976. The evolution of perma­ 1975. Perennially frozen peatlands in the western Arctic and Subarctic frost in the taiga and in the forest-tundra, western -Labrador of Canada. Canadian Journal of Sciences, 12:28-43. Peninsula. Canadian Journal of Forest Research, 6:203-220. Zoltai, S.C. and Vitt, D.H., 1990. Holocene climatic change and the distri­ Seppàlà, M., 1980. Stratigraphy of a silt-cored palsa, Atlin region, British bution of peatlands in western interior Canada. Quaternary Research. Columbia, Canada. Arctic, 33:357-365. 33:231-240.

Géographie physique e! Quaternaire, 49(1), 1995